Conventional model-based controller design approaches are based on the assumption that a perfect “nominal” model of the system is given. However, in practice, an exact system model is often either not known or is difficult to obtain. This course addresses the case where a control loop has to be robust with respect to uncertainties in the system model, that is, the control loop has to show desired behavior such as stability for a family of system models.
To this end, the course provides analysis and design techniques for the robust control. The analysis includes the formulation and verification of stability and performance criteria in the form of function norms that allow the representation of structured and unstructured model uncertainties. The controller design focuses on the H-infinity control.